Abstract
While commissioning Varian's Portal Dose Image Prediction (PDIP) algorithm for portal dosimetry, an asymmetric radial response in the portal imager due to backscatter from the support arm was observed. This asymmetric response led to differences on the order of 2%–3% for simple square fields (<20×20 cm2) when comparing the measured to predicted portal fluences. A separate problem was that discrepancies of up to 10% were seen in measured to predicted portal fluences at increasing off‐axis distance (>10 cm). We have modified suggested methods from the literature to provide a 1D correction for the off‐axis response problem which adjusts the diagonal profile used in the portal imager calibration. This inherently cannot fix the 2D problem since the PDIP algorithm assumes a radially symmetric response and will lead to some uncertainty in portal dosimetry results. Varian has recently released generic “2D correction” files with their Portal Dosimetry Pre‐configuration (PDPC) package, but no independent testing has been published. We present the comparison between QA results using the Varian correction method to results using our 1D profile correction method using the gamma passing rates with a 3%, 3 mm criterion. The average, minimum, and maximum gamma pass rates for nine fixed‐field IMRT fields at gantry 0° using our profile correction method were 98.1%, 93.7%, and 99.8%, respectively, while the results using the PDPC correction method were 98.4%, 93.1%, and 99.8%. For four RapidArc fields, the average, minimum, and maximum gamma pass rates using our correction method were 99.6%, 99.4%, and 99.9%, respectively, while the results using the PDPC correction method were 99.8%, 99.5%, and 99.9%. The average gamma pass rates for both correction methods are quite similar, but both show improvement over the uncorrected results.PACS numbers: 87.55.Qr, 87.55.N‐
Highlights
44 Hobson et al.: Comparison of corrections for portal dosimetry fields (< 20 × 20 cm2) when comparing the measured to predicted portal fluences, which has been noted by other authors.[15,16,17,18,19,20] the asymmetric response due to the backscatter from the arm is just one problem
The simplest and most practical method within the clinical system to correct for the off-axis response is the method provided by Bailey et al,(21) which is a 1D correction that adjusts the diagonal profile used in the detector calibration
We present the results for different field types using a modified version of the method used in Bailey et al[21] and compare them to results collected using Varian’s Portal Dosimetry Preconfiguration (PDPC) package in order to determine which method to use in our clinic
Summary
44 Hobson et al.: Comparison of corrections for portal dosimetry fields (< 20 × 20 cm2) when comparing the measured to predicted portal fluences, which has been noted by other authors.[15,16,17,18,19,20] the asymmetric response due to the backscatter from the arm is just one problem. The simplest and most practical method within the clinical system to correct for the off-axis response is the method provided by Bailey et al,(21) which is a 1D correction that adjusts the diagonal profile used in the detector calibration. This inherently cannot fix the 2D problem (i.e., the backscatter from the arm) since the PDIP algorithm assumes a radially symmetric response and leads to some uncertainty in portal dosimetry results. We present the results for different field types (simple square or rectangular fields, IMRT fields, and RapidArc fields) using a modified version of the method used in Bailey et al[21] and compare them to results collected using Varian’s PDPC package in order to determine which method to use in our clinic
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
